compartments the superior specific gravity of the material being removed will prevent any low grade material from the top layers in the vicinity of the pen from working their way down and underneath its lower edge. The discharge will draw from all parts of the "bed" or lower layers. In the middling compartments the slight differences of specific gravity between the top layers of tailing grade and the lower ones which it is desired to draw off is not sufficient to prevent the top pieces near the pen from being drawn down by the pull of the discharge. The discharging force is so feeble that it will pull from the top layers in the immediate vicinity of the pen rather than from points distant in the bed from the pen and there is not enough settling ratio left to overcome this tendency.
Tailings will go into the middlings and since they displace middlings the latter will go to waste. For difficult middlings-tailings separations a multi-compartment jig is required.
U ^*'« U
FIG. 3.—Harz jig.
That the Harz jig is capable of making very close separations is proved by the following experiments by the writer on material which contained but little over 0.50 per cent metallics consisting of lead and zinc and but a trifle in the way of silver reckoned as percentage. This material was jigged not for its base metal content but in hopes of finding a way of reducing the silver content. The ore consisted of gray copper, galena and zinc blende. Much of the silver was intermixed with or closely associated with the galena but there was also isolated amounts of gray copper. The ore was very clayey. In the tests material ranging in size from 1 to M in- was fed to the jig. The large range of size fed at one time makes the tests all the more remarkable. In preparing the material for the tests it was screened dry and there was much material in the form of clay balls which disintegrated in the jig and flowed away from the tanks which were assembled around the jig for catching the products made. This slime which went thus to waste will account for most of the losses which are summed up under the head of " unaccounted " in the tabulation which follows. In preliminary tests an endeavor was made to run the top discharges of the jig but the material issuing was found to differ very little from the feed in point of assay. In the test which is tabulated the test jig was fed at the rate of 1 ton an hour for periods of 20 min. The beds were then removed and the beds skimmed. The beds were then thrown back and the test was resumed. Four sub tests were so made and the summary is the result of weighing and assaying the beds so removed, the hutches or material which